JP3147859B2 - Video signal multi-channel encoder - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-channel video signal encoding apparatus, and more particularly to a multi-channel video signal encoding system for multi-channel video signals.
The present invention relates to a multi-channel encoding device that outputs a multiplexed stream after compression encoding by an EG method and multiplexing.

[0002]

2. Description of the Related Art Conventionally, a video signal such as a multi-channel moving image has been converted to an MPEG (Moving Picture Coding Experts Group).
A multi-channel encoding device that compresses and encodes data according to the p) method, multiplexes them, and outputs the multiplexed stream is known (Japanese Patent Application Laid-Open Nos. 5-111014 and 6-62).
393, JP-A-7-184192, JP-A-7-27
No. 4179).

Here, when performing multi-channel encoding of a video signal, each channel performs variable rate encoding,
In some cases, constant rate coding is performed. The multiplexing method is
Statistical multiplexing is used for variable rate coding, and deterministic multiplexing is used for constant rate coding. In statistical multiplexing, even though the code rate of each channel is variable, the multiplexing rate after statistical multiplexing is stochastically substantially constant.

[0004]

However, the fact that the multiplexing rate after the statistical multiplexing is constant means that it tends to be stochastically, and there is no guarantee that the multiplexing rate will always be constant. In this case, there is a problem that transmission failure may occur. Although deterministic multiplexing can avoid transmission failure, it requires a rate buffer to perform rate smoothing for encoding each channel at a constant rate, which leads to an increase in circuit size, and the encoded data is stored in the rate buffer. Is temporarily accumulated, so that there is also a problem that an encoding delay increases.

[0005] The present invention has been made in view of the above points,
It is an object of the present invention to provide a video signal multi-channel encoding device capable of preventing transmission failure, which is a significant service degradation in performing multi-channel two-way communication.

Another object of the present invention is to provide a multi-channel video signal encoding apparatus capable of preventing an increase in encoding delay, which is an obstacle to communication in performing multi-channel bidirectional communication. To provide.

[0007]

In order to achieve the above object, the present invention provides an image encoding unit for performing high-efficiency encoding of an input video signal according to a predetermined method and outputting encoded data, A scene change detection unit that detects a scene change, a count value decoding unit that outputs a picture type indicating an encoding processing method to be processed by the image encoding unit, an output picture type of the count value decoding unit, and the actual value of the image encoding unit. And a virtual buffer operation unit that adaptively controls the amount of generated code of the image coding unit in accordance with the generated code amount of each of the multi-channel input video signals. Output a count value sequentially and cyclically to the unit, and when a scene change detection signal is input from any of the scene change detection units existing as many as the number of channels of the input video signal, the scene change is detected. A structure determination unit for forcibly performing intra-frame encoding by a virtual buffer operation unit and an image encoding unit using a count value of a channel corresponding to a scene change detection unit that has output an image detection signal as a predetermined value, and an input video signal The coded data output in parallel from the image coding units having the same number of channels are multiplexed and output as a multiplexed stream.
If it detects that it may exceed the
Comprising a multiplexer for outputting a tea inhibit signal, I picture relevant
All virtual backups are received by the
Intra-frame encoding by the
In this configuration , prediction encoding other than intra-frame encoding is performed .

According to the present invention, a multi-rate input video signal is subjected to variable rate encoding by an image encoding section under control of a virtual buffer operation section for each channel, and
Based on the count value from the structure determining unit, the channels for performing intra-frame encoding can be efficiently arranged and multiplexed to make the multiplexed stream rate constant. In the present invention,
When a scene change is detected, the count value of the channel corresponding to the scene change detection unit that outputs the scene change detection signal is set to a predetermined value. Intra-frame encoding is possible.

Further, according to the present invention, a virtual buffer operation for adaptively controlling the generated code amount of the image encoding unit according to the output picture type of the count value decoding unit and the actual generated code amount of the image encoding unit. Since the units are provided independently for each channel of the multi-channel input video signal, the amount of code generated by the image encoding unit can be adaptively controlled for each individual channel.

The present invention further comprises means for outputting, when the multiplexing unit detects that there is a possibility that the multiplexed stream may exceed a predetermined code rate, an I picture inhibition signal. The intra-frame encoding by all the virtual buffer operation units and the image encoding units is inhibited by the structure determining unit that has received the picture inhibition signal, and the prediction encoding other than the intra-frame encoding is performed.

According to the present invention, when there is a possibility that the output multiplexed stream of the multiplexing unit may exceed a predetermined code rate,
Since all of the image encoding units perform predictive encoding other than intra-frame encoding, it is possible to prevent a predetermined code rate from exceeding a predetermined code rate by prohibiting intra-frame encoding with the largest amount of generated code.

[0012]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a multi-channel video signal encoding apparatus according to the present invention. This embodiment, operate independently from each other with respect to an input video signal of the N-channel, the image encoding unit 1 ₁ to 1 _N, the count value decoder 2 ₁ to 2 _N, the scene change detection unit 3 ₁ to 3
_N , a single GOP structure determining unit 5 and a multiplexing unit 6.

[0013] The image encoding unit ₁ 1 to 1 _N may be compression-encoded by a variable length rate based an input video signal to the MPEG scheme.
Here, in the MPEG system, as is well known,
Normally, a video signal of each screen is obtained as an encoded image of one of an I picture, a P picture, and a B picture. The I picture is an intra-frame coded image that codes all blocks in one frame in the same order as the original image, and has the largest code amount among the three pictures. The P picture is an inter-frame forward prediction coded image. Further, the B picture is a bidirectional predictive coded image obtained by predictive coding from past and future I pictures or P pictures, and has the smallest code amount among the three pictures. A picture group appropriately combining these three pictures is a group of pictures (G
OP). However, only one I picture exists in one GOP.

When the picture count value a input from the GOP structure determining section 5 matches a specific value pre-assigned to itself, the count value decoding sections 2 ₁ to 2 _N execute pictures to be subjected to intra-frame encoding. type (i.e., I picture) outputs to the corresponding virtual buffer arithmetic unit 4 ₁ to 4 _N, the virtual buffer arithmetic unit 4 when different from the predetermined value to the corresponding picture type of B picture or P picture to be a predictive coding ₁ to
To output to 4 _N.

The scene change detectors 3 _{1 to} 3 _N detect the presence or absence of a scene change in the input video signal, and supply a detection signal to the GOP structure determiner 5 when a scene change is detected. The virtual buffer operation units 4 ₁ to 4 _N are provided with a count value decoding unit 2 ₁
Based on the picture type from to 2n, after determining the amount of information generated, and outputs the determined quantization step size corresponding to the error between the actual generated information amount to the image encoding unit 1 ₁ to 1 _N.

The GOP structure determining unit 5 normally operates from 1 to N
Are sequentially and cyclically counted and output as a picture count value a for each frame period. When a scene change detection signal is input, the picture count value a is set to the detection unit number. Accordingly, the count value decoding unit corresponding to the input video signal in which the scene change is detected decodes the I picture type, and the input video signal in which the scene change is detected is
Image encoding processing is performed as a picture type.

Further, the multiplexing section 6 includes image encoding sections ₁₁ to 1 _N
And outputs a multiplexed stream when there is a high possibility that the output multiplexed stream will exceed a predetermined fixed rate. During the GOP structure determination unit 5 that are input I-picture inhibit signal, and outputs the picture count a predetermined value other than 1 to N, from the count value decoder 2 ₁ to 2 _N of P picture or B-picture Output picture type. After the cancellation, the counting is restarted from the picture count value immediately before the setting.

Next, the operation of this embodiment will be described with reference to FIG.
This will be described with reference to the block diagram of FIG. In Figure 1, an input video signal for each channel a total of N channel is provided corresponding to each channel, respectively, operate independently of each other, the image encoding unit 1 ₁ -
1 _N, the count value decoder ₂ 1 to 2 _N, are supplied to the scene change detection unit ₃ 1 to 3 _N.

The image encoding units 1 ₁ to 1 _N are connected to channels # 1 to # 1.
After separately performing high-efficiency encoding processing such as the MPEG method on the input video signal of #N, the obtained encoded stream is supplied to the multiplexing unit 6. Coding control method, the quantization step size entered from the virtual buffer arithmetic unit 4 ₁ to 4 _N per coding control unit, coding control of calculating the amount of information generated virtual buffer arithmetic unit 4 ₁ to 4 _N By supplying it to

On the other hand, the GOP structure determining unit 5 counts the number of pictures up to the number of pictures of all the channels, and
Each supplies simultaneously such picture count a shown in the count value decoder 2 ₁ to 2 _N. Count decoder 2 ₁ to 2 _N is, I-picture from the input picture count a, P and B pictures and decoded by a preset scheme of the picture type in the virtual buffer arithmetic unit 4 ₁ to 4 _N Supply each.

The count value decoder 2 _1, picture count a
Is "1", picture type # 1 (f) is set as an I picture. For values of 2 to N other than the picture count value a of 1, the picture type # 1 (f) is changed to P picture or B picture.
Let it be a picture. The relationship between the P picture and the B picture can be set arbitrarily, but usually the same number of B pictures are allocated between the P picture and the P picture or between the P picture and the I picture. FIG. 2B shows the above picture type # 1 (f).

The count value decoding unit 2 _N sets the picture type #N (k) as an I picture only when the picture count value a is N. As for the picture count value a of 1 to N-1 other than N, the picture type #N (k) is changed to P picture or B picture.
Let it be a picture. The relationship between the P picture and the B picture can be set arbitrarily, but usually the same number of B pictures are allocated between the P picture and the P picture or between the P picture and the I picture. FIG. 2C shows the above picture type #N (k). Note that each count value decoding unit may be set so that a picture type obtained by delaying picture type # 1 (f) by N pictures becomes picture type #N (k).

Virtual buffer operation unit 4₁~ 4_NEach of
Is a count value decoding unit 2₁~ 2_NOf the total
Based on the picture type given by the numerical decoding unit,
Is determined, and the image encoding unit 1 ₁~ 1_NHorse
Actual occurrences sequentially input from the corresponding image coding unit
Compare the quantization step size with the corresponding image
This is given to the image encoding unit. Thus, for example, a virtual buffer
Arithmetic part 4_NIs the count value decoding unit 2_NPicture tie given by
The target amount of generated information is determined based on
Image encoding unit 1_NOf actual generated information #N input from
The quantization step size #N (h) is
Image encoding unit 1_NTo supply.

[0024] In addition, a scene change detection unit 3 ₁ ~3 _N is,
For example, when the level difference between the input video signal of the previous frame and the current frame is equal to or larger than a preset value, it is detected as a scene change, and a scene change detection signal is output to the GOP structure determination unit 5, respectively. GOP structure determination unit 5
Is, x-th (x is any one value among the 1 to N) when the scene change detection signal from the scene change detection unit 3 _x of is inputted, the count value decoding corresponding to the scene change detection unit 3 _x The value of the picture count value a of the unit 2 _x is forcibly set to x.

[0025] Thus, for example, GOP structure determination unit 5, if such shown in FIG. 2 (D) from the scene change detection unit 3 _N scene change detection signal #N (n) is input, the scene change detection unit 3 _N the value of the picture count a corresponding count value decoder 2 _N to be set to force the N as shown in FIG. 2 (a). Thus, when a scene change occurs, the image encoding unit that encodes the input video signal in which the scene change has occurred preferentially prefers the GOP structure to the IOP.
The control to make a picture is performed.

The multiplexer 6 outputs a multiplexed stream by performing multiple processing encoded stream input from the image encoding unit 1 ₁ to 1 _N. At this time, when there is a high possibility that the multiplexed stream to be output exceeds the predetermined coding rate, the multiplexing unit 6 supplies an I picture inhibition signal to the GOP structure determining unit 5. GOP structure determination unit 5, while the I-picture disable signal is input, the picture count a as a value other than 1 to N (e.g. 0), the I picture at all count decoder 2 ₁ to 2 _N The decoding of the picture type is prohibited.

Therefore, for example, when the picture inhibit signal is output from the multiplexing unit 6 at the timing shown in FIG. 2E, the picture count value a becomes 0 as shown in FIG. , The count value decoding units 2 ₁ and 2 _N correspond to FIG.
As shown in (C), a picture type other than the I picture is output. Each of the image encoding units ₁₁ to 1 _N thus obtained
Are multiplexed by the multiplexing unit 6 and output as a multiplexed stream at a constant rate.

As described above, in this embodiment, by shifting the phase of the GOP structure of the video signal between the respective channels, the encoding of the I picture having a large amount of generated information is dispersed on the time axis and the multi-channel multiplexing is performed. The target generated information amount after the conversion is fixed. That is, while performing the respective variable rate coding separately by the image encoding unit 1 ₁ to 1 _N of the input video signal of the N-channel, after multiplexing the I-picture generation amount information is biased to efficiently disposed between the channels Since the multiplexed stream is set at a constant rate, a rate buffer is not required, and an increase in delay time can be prevented.

Further, in this embodiment, the deviation between the actual generated code amount after the target amount of information generated above and coding, achieving corrected by feedback control using a virtual buffer arithmetic unit 4 ₁ to 4 _N However, since the image coding is performed, variable rate coding in which the rate after multiplexing becomes constant can be realized.

If the predetermined coding rate is about to be exceeded, I pictures are prohibited to reduce the amount of generated codes. In addition, scene change images reduce the prediction efficiency in the time direction between pictures (between frames or fields),
GOP structure phase control is performed such that an image in which a scene change is detected is an I picture that preferentially suppresses only the spatial redundancy in the picture. That is, when a scene change occurs, the input video signal is forcibly encoded with an I picture, so that the predictive encoding efficiency between pictures can be reduced.

Furthermore, likely exceeding since to carry out the coding control is provided a virtual buffer arithmetic unit 4 ₁ to 4 _N individual channel unit it is stable control, the set value rate of the multiplexed stream In such a case, encoding of the I-picture having the largest code amount among the three picture types is prohibited, so that the rate of the multiplexed stream can be stabilized, and transmission failure can be prevented.

[0032]

As described above, according to the present invention,
The multi-channel input video signal is variable-rate coded for each channel, and the channels for intra-frame coding are arranged efficiently to keep the multiplexed stream rate constant. Therefore, it is possible to eliminate the need for a rate buffer, and to prevent an increase in coding delay which is a problem in achieving communication in two-way communication.

Further, according to the present invention, when a scene change is detected, a scene change occurs by setting the count value of the channel corresponding to the scene change detection unit that outputs the scene change detection signal to a predetermined value. Since the encoding of the input video signal is forcibly performed in the intra-frame, the prediction encoding efficiency between pictures can be reduced.

Further, according to the present invention, the amount of generated code of the image coding unit is adaptively controlled for each channel, so that even if the input video signal of each channel is variable-rate coded, a stable code amount can be obtained. The generated code amount can be controlled, and the rate of the multiplexed stream can be made constant.

Still further, according to the present invention, when there is a possibility that the output multiplexed stream of the multiplexing unit may exceed the predetermined code rate, all the image coding units perform predictive coding other than intra-frame coding. Is performed, the intra-frame encoding with the largest generated code amount is prohibited so that the predetermined code rate is not exceeded, so that the system can be stabilized and transmission failure can be prevented.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention.

FIG. 2 is a time chart for explaining the operation of FIG. 1;

[Explanation of symbols]

1 ₁ to 1 _N Image encoding unit 2 ₁ to 2 _N Count value decoding unit 3 _{1 to} 3 _N Scene change detection unit 4 ₁ to 4 _N Virtual buffer operation unit 5 GOP structure determination unit 6 Multiplexing unit

Claims (5)

(57) [Claims] 1. An image encoding unit that performs high-efficiency encoding of an input video signal according to a predetermined method and outputs encoded data; a scene change detection unit that detects a scene change of the input video signal; A count value decoding unit that outputs a picture type indicating an encoding processing method to be processed by the encoding unit, and an image according to an output picture type of the count value decoding unit and an actual generated code amount of the image encoding unit. A virtual buffer operation unit for adaptively controlling the amount of generated code of the encoding unit is provided independently for each channel of the multi-channel input video signal; And when a scene change detection signal is input from any of the scene change detection units that exist for the number of channels of the input video signal, the scene change detection is performed. A structure determination unit for forcibly performing intra-frame encoding by the virtual buffer operation unit and the image encoding unit using the count value of the channel corresponding to the scene change detection unit that has output the signal as the predetermined value; and outputs as a multiplexed stream encoding data outputted in parallel from the image encoding unit that exist only the number of channels of signals are multiplexed, the multiple Kas
Trim may exceed a given code rate
When detecting, the structure which comprises a multiplexer for outputting the I-picture disable signal, received the I picture inhibiting signal
All of the virtual buffer operation units and image
Intra-frame encoding by the image encoding unit is prohibited and the frame
A video signal multi-channel encoding device for performing predictive encoding other than intra- system encoding. 2. When the number of channels of the input video signal is N, the structure determining unit sequentially and cyclically outputs count values from 1 to N to the N count value decoding units. 2. The multi-channel video signal encoding apparatus according to claim 1, wherein: 3. The virtual buffer operation unit corresponding to a picture type for which intra-frame encoding is performed when the input count value matches a unique predetermined value pre-assigned to itself. 3. The video signal according to claim 1, wherein when the value is different from the predetermined value, a picture type for performing predictive encoding other than the intra-frame encoding is output to the corresponding virtual buffer operation unit. 4. Multi-channel encoding device. 4. The virtual buffer operation unit determines a target generated code amount based on an output picture type of the count value decoding unit, and determines an actual generated code amount sequentially input from the image coding unit. 2. The video signal channel encoding apparatus according to claim 1, wherein a comparison is made, and a quantization step size of the image encoding unit is controlled according to the comparison result. 5. The structure determination unit according to claim 1, wherein the structure determination unit is a GO of the MPEG system.
The P value structure is determined, and the count value decoding unit determines the I picture, P
The multi-channel video signal encoding apparatus according to any one of claims 1 to 4 , wherein the apparatus outputs one of a picture type of a picture and a B picture.